Distinct efficacy of postoperative radiotherapy in patients with resected pathological Stage IIIA-N2 lung squamous cell carcinoma

Background. The beneficial effect of postoperative radiotherapy (PORT) on resected pathological IIIA-N2 (pIIIA-N2) non-small cell lung cancer (NSCLC) has been a subject of interest with controversy. The aim of this study was to distinguish the clinical efficacy of PORT on lung adenocarcinoma (LADC) and squamous cell carcinoma (LSCC) among pIIIA-N2 NSCLC. Methods. Between October 2010 and September 2016, 288 consecutive patients with resected pIIIA-N2 NSCLC at Beijing Chest Hospital were retrospectively analyzed. There were 194 cases of adenocarcinoma (ADC), 85 cases of squamous cell carcinoma (SCC), 5 cases of large cell carcinoma, 3 cases of adenosquamous carcinoma, and 1 case of clear cell carcinoma. In pIIIA-N2 LADC and LSCC 42 and 19 cases received PORT, respectively. Life Table was used for univariable analyses of factors affecting the rate of overall survival (OS), locoregional recurrence-free survival (LRFS) and distant metastasis-free survival (DMFS). Multivariable Cox proportional hazard models were used to evaluate risk factors affecting OS, LRFS and DMFS. Results. In 194 cases of pIIIA-N2 LADC, smoking index (SI) < 400 (p = .000), a lower number of positive nodes (p = .009), a single N2 station (p = .012) and treatment with postoperative adjuvant chemotherapy (POCT) (p = .006) were independent prognostic factors for OS in multivariable analyses. Other beneficial factors included the use of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) or anaplastic lymphoma kinase (ALK) inhibitors in univariable analyses. PORT failed to show benefit in prolonging the OS of pIIIA-N2 LADC patients. In contrast, in 85 cases of pIIIA-N2 LSCC, PORT alone was a significant positive prognostic factor for OS (p = .027), target volume; LRF: local-regional failure; DM: distant metastasis; LRFS: locoregional recurrence-free survival; DMFS: distant metastasis-free survival; OS: overall survival; MT: median time; MST: median survival time; DRFS: distant recurrence-free survival; EML4: microtubule-associated protein-like 4.


Introduction
Lung cancer is the leading cause of cancer related deaths worldwide [1]. Non-small cell lung cancer (NSCLC) accounts for 80-85% of all cases of lung cancer. Surgery remains the most important treatment for stage I/II and IIIA NSCLC. In particular, patients with IIIA-N2 disease require surgery and supplementary treatments to improve the prognosis [2]. However, after adjuvant chemotherapy, this patient population still had a high loco-regional recurrence rate of up to 40% [3,4]. Therefore, postoperative radiotherapy (PORT) has been incorporated into multidisciplinary management in hope that it may reduce local-regional recurrences. Nevertheless, due to lack of phase III randomized clinical trials (RCTs) [5,6] to evaluate the effect of PORT on resected pathological IIIA-N2 (pIIIA-N2) NSCLC, its benefit remains unclear and even controversial. In many retrospective studies [7][8][9], the efficacy of PORT was analyzed under the general population of pIIIA-N2 NSCLC (without separation of histologic subtypes), that may undermine the benefit for certain patient subpopulations. Lung adenocarcinoma (LADC) and squamous cell carcinoma (LSCC) are the most frequent histologic subtypes of NSCLC, accounting for 50% and 30% of all cases, respectively [10]. During the past decades, LADC had received treatments consisting of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) and anaplastic lymphoma kinase (ALK) inhibitors with improved prognosis [11]. Unfortunately these inhibitors did not show the therapeutic effect on LSCC [12,13].
Since LSCC had a higher local recurrent rate (21% vs. 14%) and lower distant metastatic rate (7% vs. 11%) than LADC in patients with resected NSCLC [14] , more rigorous local treatment using PORT to eradicate micro-residues of the tumor may achieve improvement in OS in patients with LSCC [9].
These observations prompted us to analyze stage pIIIA-N2 LADC and LSCC during the period of Patients survived 4 months after radical resection in Beijing Chest Hospital were enrolled in this study.
The medical records and follow-up data of the patients were retrospectively analyzed, including: gender, age, smoking index, histology, pathological TNM stage, types of surgery, types of N2 (N2a1 N2a2, N2b), number of positive nodes, percentage of positive nodes, number of N2 stations, postoperative adjuvant chemotherapy (POCT), PORT, postoperative adjuvant EGFR TKIs or ALK inhibitors, patterns and times of recurrence, salvage treatments, and survival status.

Surgery
The surgical methods of 288 patients were divided into thoracic surgery (261 cases) and thoracoscopic surgery (27 cases). Types of surgery contained single lobectomy (212 cases), compound lobectomy (18 cases), sleeve resection (12 cases), and total lung resection (pneumonectomy, 46 cases). Single lobectomy, compound lobectomy and sleeve resection were classified as lobectomy in data analysis. Complete mediastinal lymph node dissection or systematic mediastinal lymph node sampling was performed during surgery, with an average removal of 20 nodes per patient.

Chemotherapy
POCT was administered with a cisplatin-or carboplatin-based regimen, with a median of four cycles.
A total of 38 patients did not receive POCT due to asthenia, refusal, or choice by the physician.

Therapy with EGFR TKIs or ALK inhibitors
Among 194 cases of LADC, 75 patients received therapy with EGFR TKIs or ALK inhibitors (5 patients with ALK inhibitors and 70 with EGFR TKIs). The inhibitors were given to 8 patients as adjuvant therapy after surgery and to 67 patients after tumor relapse.

Radiotherapy (PORT)
PORT was performed in 42 of 194 LADC patients and 19 of 85 LSCC patients. The administration of PORT was based on the radiation oncologists' decision or surgeon's referral. Extensive mediastinal lymph node involvement was the main indication for PORT. Techniques included three-dimensional conformal radiotherapy (3D-CRT, 21 cases) and intensity modulated radiotherapy (IMRT, 40 cases).
Clinical target volume (CTV) included surgical margin, ipsilateral hilum, and high-risk ipsilateral mediastinal drainage lymph area. The planning target volume (PTV) was defined as the CTV plus 0.5-0.8 cm margins. The therapies were administered with a linear accelerator using 6-8 MV x-ray at 180-200 cGy per fraction, 5 days per week, to an average total radiation dose of 5918 cGy. PORT was used for an average of 4.38 months after surgery.

Follow-Up
The patients were followed up every 3 months after surgery for the first 2 years and every 6-   Table 1 and Table 2. Analyses of 85 LSCC patients are summarized in Table 3 and Table 4.  Multivariable Cox proportional hazard models (Forward: LR) were used to adjust risk factor distributions between patient groups. A statistically significant difference was set as p < .05, represented by "*". *: p < .05. **: p < .01. Abbreviations: POCT = postoperative chemotherapy, HR = hazard ratio, CI = confidence interval.  Multivariable Cox proportional hazard models (Forward: LR) were used to adjust risk factor distributions between patient groups. A statistically significant difference was set at p < .05, represented by "*". *: p < .05. **: p < .01. Abbreviations: PORT = postoperative radiotherapy, HR = hazard ratio, CI = confidence interval.
Stratified analysis was used to analyze 85 LSCC patients who were further divided into two groups. 19 received PORT and 66 were in the non-PORT group. Table 5 compared characteristics of patients in the PORT and non-PORT groups. (p = .005), a lower number of positive nodes (p = .001), a lower percentage of positive nodes (p = .003), a single N2 station (p = .002) and receiving therapy with EGFR TKIs or ALK inhibitors (p = .017). Patient gender, age, types of surgery, pathologic T stage, and treatment with PORT or with POCT were not prognostic factors for OS. In multivariate analysis, SI < 400 (p = .000), a lower number of positive nodes (p = .009) and a single N2 station (p = .012) were independent prognostic factors for OS. Treatment with POCT was an independent prognostic factor both for OS (p = .006) and LRFS (p = .032) (Fig. 1). Other factors were not significantly associated with OS. Also, PORT was the only independent favorable prognostic factor for LRFS (p = .014) and DMFS (p = .030) (Fig. 2).
In order to define types of patients who benefited more from PORT, stratified analysis was used in Table 5. Life Table analysis showed that patients with older age (≥ 65 years) (p = .031) and a lower T stage (p = .042) in PORT group had longer OS. Male (p = .090), SI < 400 (p = .098), lobectomy (p = .095), treatment without POCT (p = .072) and a higher number of positive nodes (p = .071) in patients of PORT group had no statistically significant difference compared with no PORT group.
Therefore, PORT demonstrates a unique favorable prognosis value for pIIIA-N2 LSCC.

Discussion
Traditionally, although both were categorized by NSCLC, we regard LADC and LSCC as different cancers due to their distinct cells of origin, unique molecular characteristics and dissimilar clinical responses to treatment. LSCC typically originates from bronchial epithelium of larger and more proximal airways (basal cells), mostly from central lung and more closely associated with smoking and chronic inflammation [15,16]. Patients with LSCC trend to have a lower distant recurrence-free survival (DRFS) rate than those with LADC [17]. Therefore, the main objective of the therapies for LSCC patients with a high risk of local recurrence is to eradicate potentially residual microscopic tumors with surgery, such as at the resection margin or in mediastinal node areas. In a randomized study, of 366 patients with resected pN1-N2 NSCLC [18], PORT resulted in a significantly lower locoregional recurrence rate. PORT may reduce local-regional recurrence, and was more effective for patients of resected pIIIA-N2 LSCC [19]. In a randomized study of 773 lung cancer assigned 230 patients with resected stage II or stage III LSCC have received either PORT or no adjuvant treatment.
The overall recurrence rate was significantly lower with PORT in patients bearing N2 disease [20].
Also, in LSCC, driver oncogenic alterations such as EGFR [11] and ALK [12] gene rearrangements are rarely detected, and these tumors are generally chemotherapy-insensitive [21]. Our study demonstrates that pIIIA-N2 LSCC patients treated with PORT had a longer median LRFS time with prolonged OS. PORT patients in our study were randomly selected to minimize the inconsistencies indications. Therefore, the advantage of PORT for pIIIA-N2 LSCC was more convincingly illustrated.
PORT also indicating a higher sensitivity of squamous cell carcinoma port in another anatomic location significantly benefit for esophagus squamous cell carcinoma after surgery [22].
LADCs are thought to originate from the bronchiolar or alveolar epithelium (Clara cells or type II pneumocytes), mainly located in the peripherally located smaller airways with glandular histology features and biomarkers consistent with an origin in the distal lung [15,16]. The discovery of EGFR gene mutations and echinoderm microtubule-associated protein-like 4-ALK(EML4-ALK) gene rearrangement led to the development of targeted therapy [23,24], which improved clinical outcomes in a subset of LADC patients. In our study 194 LADC cases showed improved OS by POCT (p = .006), and subsequent therapy with EGFR TKIs or ALK inhibitors further prolonged OS (p = .017) of patients with tumor recurrence. The MST was 54.77 months of pIIIA-N2 LADC patients, including that addition of EGFR-TKIs to POCT are able to prolong the OS for resected LADC among NSCLC [25].
Based on the superior disease-free survival, reduced toxicity, and improved quality of patient life, adjuvant EGFR-TKIs therapy is an ideal option for II-IIIA (N1-N2) EGFR-mutant LADC among NSCLC [26].
The results derived from our study clearly distinguish pIIIA-N2 LSCC from LADC in terms of their tissue origination, etiology, genetic characteristics and most importantly, their sensitivity to PORT. Based on insensitivity of pIIIA-N2 LADC to PORT, we consider it no-longer appropriate to collectively categorize LADC and LSCC into a vague term of NSCLC.

Conclusion
We propose that pIIIA-N2 LSCC should be singled out into a unique patient population who